240 V Transistors and chips

[Chris Holmes]

Hi All,

I recently bought and installed a Manrose timer board for one of their inline fans (so it runs on after the lights are turned off). I installed it this weekend (actually driving a similar fan from another manufacturer).

It's working fine, but I found it a little curious...?.

It's a tiny PCB with quite small tracks. It only has about 10-12 components, from memory, a small transistor at the input end together with a pot to adjust the delay, a big resistor, a handful of tiny things most of which are probably resistors, but one or two could be diodes or something else a chip, and another little transistor at the output end. The permenant live is just a track that goes straight to the output, and the neutral goes most of the way there and then gets switched to the output.

Thing is, in my limited experience, transistors and chips required a power supply of some sort to convert mains to 5V and or 12V DC. But I couldn't see anything on the board that I recognised as being capable of doing this. Are there now 240V AC chips? Have there always been and I've just not come across them before?

I got it for about £10 by the way (on special offer, reduced by 50-ish %), so a nice cheap simple way to add a timer to something that doesn't require much wattage.

[Roger Hayter]

I think the secret is low current consumption. To drop (most of) 240V at, say, 5mA needs only to dissipate 240 x 5 x 10^-3, that is about 1.2W and the large resistor can probably do that even air cooled. Only few low power components are needed to voltage limit, smooth and regulate the low voltage supply. There should be an electrolytic capacitor though, unless anyone can give a clue how to avoid it.

[Dave Liquorice]

Agreed but I think "capacitive droppers" are more effcient.

Good regulation probably isn't needed as the load will be known and thus the voltage determined with component values.

A capacitive dropper uses a non-polarised capacitor IIRC but normally a wire ended device rather than surface mount.

[Martin Brown]

Chances are that as a circuit that doesn't have to be exposed to people they use a capacitive divider and a diode to generate the small amount of DC power that the timer circuit needs to operate and that the thing doing the switching is a power thyristor or triac.

Those have been available at mains voltages almost forever.

[Fredxx]

The small transistor is likely to be a Thyrister or SCR.

There are chips that can regulate from 240V to 5V using the first part of the AC waveform though never used one. Mains voltage capable chips are available but not common place.

A resistor, diode and zener can get you a low power 5V.

[Paul]

Plugged "Manrose timer board" into Google and got an iFixit link with a Youtube reference.

"How to Fix an Extractor Fan Timer"

CD4001BE, big-ass RC (analog) timer circuit ? Like making an amplifier out of CMOS logic.

Have a look and see if it's similar to yours.

There is a large resistor that does the dropping to bring the 240V down to some lower level. Notice in the video that his resistor is cracked or has some surface damage of some sort. Looks safe and reliable.

A non-isolated circuit ? What's not to like ?

I was hoping for a schematic, but that's too much to expect.

Paul

[Chris Holmes]

On Monday, September 28, 2020 at 11:05:01 AM UTC+1, Roger Hayter wrote: <Snip>

You are quite right Roger, I had forgotten there was a couple of small electrolytics near the input).

And thanks everyone, I shall probably never build anything like this, but it's interesting to know how they have powered what I presume is low voltage stuff.

[The Natural Philosopher]

1/. Yes there are. But probably you only have one on the board. To switch the mains. 2/. As far as the rest of the circuit is concerned the easiest way to drop voltage if the current draw is constant is via a mains rated capacitor. Then rectify and smooth the output enough. 3/ a more expensive way is to have a high frequency mains transistor chop the mains and feed that through a tiny transformer, but it sounds like that is more complex than your circuit's designer wanted to pay for..

Digital crap is so cheap that it probably uses some sort of timer, and doesn't therefore need any greatly smoothed power supply.

[The Natural Philosopher]

don't even need that. A capacitor of as it happens about 0.63µF will deliver 5mA into a small circuit and be more reliable cooler and cheaper. Feed that into a diode bridge and, if my back of envelope is right, 50µF and you will have less than 1V of ripple at 5mA.

Only few low power

well yes, and that's by going v high frequency, but then you need a choke or transformer instead...

[Dave Plowman (News]

Think dimmers. Which have been around for a long long time. Without a low volts transformer. And some later ones have quite sophisticated electronics as well as a triac or whatever.

[Jeff Layman]

Not low power - you are dropping around 230V through that resistor. Even at only 5mA that is 1.65W to be dissipated, and that is considerably more than a standard resistor will handle . Interestingly, Paul's post notes "There is a large resistor that does the dropping to bring the

240V down to some lower level."

[Harry Bloomfield, Esq.]

Chris Holmes was thinking very hard :

It will use a triac for switching and likely a mains rated capacitor and diode as a capacitive dropper, to reduce the mains voltage down to a useable AC voltage.

[Roger Hayter]

Maybe so, but I took the OP's description of the module to exclude the presence of a mains-rated capacitor.

[Fredxx]

It would be less than 1.2W but I take your point. There could also be a series cap to drop some of the volts.

Hardly an ornate solution but cheap.

[David]

I have replaced (removed) a couple of these boards over the years. They had failed, and the big resistor was quite charred in both cases!

[Dave W]

I followed your tracks and found a better video. The PCB in yours was marked 1998, but the one in mine is 2008, and has a big copper PCB land on one leg of the big resistor in an attempt to add more cooling.

(The presenter sounds like an aircraft pilot welcoming passengers).

I've worked out the circuit and added an explanation. See:

I am guessing that the IC runs on 12V and the trimmer is 2M to give a delay of 1 to 20 minutes.

[alan_m]

[Paul]

Neat. That looks about right. Fast attack, slow decay circuit. The diode in the upper left could be a 1N4007. For the price, probably not that much more for a higher rating. I think they're the same diode, just some are tested to higher voltages.

Does this mean the triac only conducts on half-cycles ???

The only hobby triac circuits I've had here, they were fired by a diac.

Paul

[Dave W]

I forgot to include a link to the video that I used for source material:

Triacs conduct in both directions when the gate is held +ve, so give full-cycle operation.

I don't know what type of 1Nxxxx diode was used - I originally put

1N4007, but changed it to 1N4004 assuming the board is made to minimal cost. I was puzzled by the component numberings, but concluded that they refer to the bins that the manual assemblers get the components from, so two diodes have the same number.

[charles]

er, no. Triacs are diodes. To get full cycle conduction you need a pair of triacs.